Rapid Microanalysis-Based Process for Nuclear Material Characterization for Nuclear Forensics

When a nuclear device is detonated or when illicit nuclear material is intercepted, the resulting debris is a chaotic mixture of melted metal, soil, and radioactive particles. To understand what happened, where it came from, and who made it, scientists must analyze this debris. This field, known as nuclear forensics, relies on measuring the specific types of atoms within the material. Traditional methods involve taking a large chunk of the sample, dissolving it entirely in acid, and running it through a series of chemical baths to separate the elements. While this approach is reliable, it is slow and treats the entire sample as a single, uniform substance. The problem is that nuclear debris is rarely uniform; it is a collection of individual particles, each potentially holding a different story about the event that created them. If a scientist dissolves the whole sample at once, these unique differences are lost, blended into a single average that hides the truth.

A team of researchers at the Pacific Northwest National Laboratory has developed a new way to look at this problem, one that focuses on the individual particles rather than the whole pile. Instead of dissolving everything immediately, their new process acts like a high-tech searchlight, scanning a sample to find the specific radioactive specks that matter most. Once these "hot spots" are located, the scientists extract just those tiny particles, dissolve them in a drop of liquid, and analyze them separately. This approach allows them to see the distinct chemical fingerprints of individual particles, revealing variations that would be invisible in a standard bulk analysis. By targeting only the most interesting pieces and using automated systems to handle the tiny amounts of liquid, the team has created a pathway that could deliver critical answers much faster than current methods allow.

The researchers tested this new workflow using two different types of simulated nuclear debris. In the first test, they used particles about one millimeter in diameter, roughly the size of a grain of sand, which contained weapons-grade plutonium and fission products. In the second, they used much smaller particles, about fifty micrometers in diameter, scattered across a filter paper. The first step in their process was to find these particles. They used a specialized imaging system called an ionizing-radiation quantum imaging detector, which acts like a camera sensitive to radiation. This device took a picture of the sample, showing exactly where the radioactive particles were located. For the larger particles, the system identified them clearly against a blank background. For the smaller particles on the filter, the system mapped out six specific areas of interest, distinguishing between spots with high radiation and spots with only background levels.

Once the locations were identified, the team moved to the extraction phase. For the larger particles, they physically picked them out and placed them on a small metal disk. For the smaller particles on the filter, they used a tiny hole punch to cut out the specific sections of the paper that contained the radiation. These tiny samples were then dissolved using a robotic liquid handler that added precise amounts of acid and heat. The goal was to get the solid particles into a liquid form using the smallest possible amount of reagents. After dissolving, the liquid was passed through custom-made, 3D-printed columns that acted as tiny chemical filters. These columns separated the different elements, isolating the uranium, plutonium, and neptunium from the other debris. Finally, the purified liquids were fed directly into a mass spectrometer, a machine that weighs atoms to determine exactly which isotopes were present and in what quantities.

The results of these experiments showed that the new method works and, more importantly, that it reveals details that the old method would miss. When the team analyzed the individual large particles, they found that while most had similar ratios of plutonium isotopes, one particle was different. It contained a higher amount of a specific plutonium isotope and a different mix of neptunium, suggesting it came from a slightly different source or process than the others. If these particles had been mixed together and dissolved as a single bulk sample, this unique particle would have been averaged out, and its distinct signature would have been lost. Similarly, when they analyzed the punched-out sections of the filter paper, they found that the uranium isotopic composition varied from one spot to another. Some spots contained almost no uranium, while others had significant amounts, and the ratio of uranium isotopes changed across the different samples. This variation suggested that the particles on the filter were not uniform, likely due to different processing activities in the area where the sample was collected.

The researchers also demonstrated that this micro-scale approach can be fast. While dissolving a large bulk sample can take anywhere from twelve to twenty-four hours or more, depending on its size and composition, their targeted micro-analysis process can be completed in a matter of hours for a small set of samples. The team integrated automated steps, from the robotic dissolution to the chemical separation, to minimize human handling and speed up the timeline. They also developed computer models to predict the best chemical steps to take based on early data, aiming to remove unnecessary steps and further reduce the time needed to get results. Additionally, they used machine learning algorithms to automatically identify the radioactive elements in the samples by analyzing the energy signatures of the radiation, a task that usually requires a human expert to review complex graphs.

This work does not replace the traditional, robust methods used in nuclear forensics but offers a complementary path for rapid initial analysis. By focusing on the micro-scale, the researchers can quickly identify the most important particles and get preliminary data on isotopic ratios and fission products without waiting for the entire bulk sample to be processed. The study confirms that nuclear debris is heterogeneous, meaning it is made of diverse parts with different histories, and that looking at these parts individually provides a clearer picture of the event. The team plans to continue refining these methods, testing them on more complex environmental samples, and integrating the automated decision-making tools to create a fully streamlined system for future nuclear forensic investigations.